The principle of microscopic reversibility at equilibrium is often invoked in the context of the feasibility of postulated kinetic mechanisms. In this context, it is argued that the De Donder relation A upsilon greater than or equal to 0, relating the affinity A and the net rate upsilon of a chemical reaction, appears to be particularly incisive, as it clearly applies to nonequilibrium systems, at the steady state or the quasi-steady state. In particular, when reactions or elementary steps are involved in closed cycles, application of the De Donder relation leads to informative restrictions concerning postulated mechanisms.